This application proposes a structural and functional study by 3-D cryo-electron microscopy (cryo-EM) into the heterodimeric kinesin-2 motor domains of mouse KIF3AB and KIF3AC to define their mechanochemistry, intramolecular communication, and structural relationship with the microtubule lattice. Unlike for many other kinesins there is still comparably little data available on kinesin-2, biochemical as well as structural. Here we would like to investigate why these kinesins employ a heterodimeric and modular motor domain, and how that feature distinguishes them functionally from other kinesins. Furthermore, we would like to compare kinesin-2 to other heterodimeric motors, Kar3Vik1 and Kar3Cik1 that we have studied extensively in the past. We will test the hypothesis that the MT binding pattern of kinesin-2 heterodimers shares common structural and functional features to heterodimeric Kinesin-14 family members such as Kar3Vik1 or Kar3Cik1, but also to kinesin-1 due to their anterograde directionality. This project will continue a long-standing collaboration between the lab of Susan Gilbert and the P.I. that dates back to 1999, much before the P.I.'s arrival at the University of Colorado at Boulder in 2006. One of the most important technical challenges of this proposal will be analyzing 3-D volumes of microtubule-bound, heterodimeric kinesin-2 motors and unambiguously identify which each of the two heads. This is substantially more complex than working with monomeric head constructs. We typically find monomeric constructs to bind to microtubules with a stoichiometry of one head per tubulin heterodimer. Hence, with a helical microtubule template (i.e. a 15-protofilament microtubule) monomeric kinesin- microtubule complexes adapt that helical symmetry and can be reconstructed accordingly. Here we will employ cryo-electron tomography (cryo-ET) 3-D reconstruction followed by statistical classification and averaging of sub-volumes extracted from tomograms. Within specific aim-1 we will apply and refine labeling strategies for kinesin-II motor head domains KIF3A, KIF3B, and KIF3C with clonable, electron- dense labels that will allow for an unambiguous separation of the heads by cryo-electron microscopy on motor-microtubule complexes. The two heads are most likely structurally too similar to be distinguished at ~1-2 nm resolution (currently about the limit that we can achieve with cryo-ET) from their shape alone. In specific aim-2 we will investigate by cryo-EM and cryo-ET the 3-D configuration of heterodimeric, tagged and native kinesin-2 motor head constructs under various nucleotide-binding conditions when interacting with microtubules. To this end we will employ various 3-D reconstruction and analysis methods that may be suitable for the predicted conditions. Finally, specific aim-3 will employ high-resolution surface shadowing to gain a very direct and unobstructed view on the interaction patterns of kinesin-2 heterodimeric motor constructs with the surface of microtubules.